PROJECT SUMMARY Recent studies revealed that the human genome encodes thousands of lncRNAs with little proteincoding capacity. LncRNAs were shown to play important roles in cancer and are potentially a new class of therapeutic targets for cancer. However, the function of the vast majority of lncRNAs in cancer remains unknown. LncRNA function often depends on its physical interactions with protein complexes. They can also influence the abundance of other mRNAs that are targeted by the same microRNAs by competing for microRNA binding, i.e., serving as competing endogenous RNA (ceRNA). Advances in genomic technologies, especially those based on next generation sequencing (NGS), provide unparalleled opportunities to characterize the functional networks of lncRNA in cancer. However, analysis and integration of different types of genomic datasets to generate testable hypotheses is challenging, and systematic approaches to characterize lncRNA function in cancer are lacking. This application describes the development of computational methods and integrative genomic strategies for systematically dissecting the functional network of lncRNA in cancer, and a combination of computational and experimental approaches to unravel several important functional networks of lncRNA in prostate cancer. Specifically, it will (1) develop a computational method for repurposing the publically available array-based data to interrogate lncRNA expression in tumor samples and utilize an integrative genomic strategy to predict lncRNAs that may be important for tumorigenesis/tumor suppression in prostate cancer via analysis of lncRNA expression profiles, clinical information and somatic genomic alteration profiles of tumor samples, (2) identify the lncRNAs that are associated with EZH2 or direct transcriptional targets of EZH2 repression that are important for prostate tumorigenesis or tumor suppression, and (3) identify the ceRNAs of AR and PTEN that mediate prostate tumorigenesis or tumor suppression. In addition to its scientific proposal, this application proposes a comprehensive training program for preparing an independent investigator in the fields of computational genomics, noncoding RNA and cancer, who develops cutting-edge computational methods, and uses a combination of computational and experimental approaches to understand structure-function relationship of noncoding RNA and the function of noncoding RNA and RNA-protein interaction in cancer. While the candidate of this application has received extensive training in biophysics, statistics, machine learning and computational genomics, this career development award will allow him to develop his experimental skills, especially those next-generation sequencing-based techniques and molecular biology experiments in human cell lines. Dr. Liu, Professor of Biostatistics and Computational Biology and Dr. Brown, Professor of Medicine will mentor the candidate in the excellent training environment of Dana Farber Cancer Institute, a part of Harvard Medical School community. A committee of experienced computational and cancer biologists will also advise him on both scientific research and career development.